Volume 22, Issue 8, Pages 2107-2117 (February 2018) Suppression of Tcf1 by Inflammatory Cytokines Facilitates Effector CD8 T Cell Differentiation  Maxime Danilo, Vijaykumar Chennupati, Joana Gomes Silva, Stefanie Siegert, Werner Held  Cell Reports  Volume 22, Issue 8, Pages 2107-2117 (February 2018) DOI: 10.1016/j.celrep.2018.01.072 Copyright © 2018 The Author(s) Terms and Conditions

Cell Reports 2018 22, 2107-2117DOI: (10.1016/j.celrep.2018.01.072) Copyright © 2018 The Author(s) Terms and Conditions

Figure 1 Tcf1 Downregulation Is Mediated by Systemic Inflammation (A) Experimental approach. P14 chimeric mice (i.e., B6.CD45.1 mice adoptively transferred with 104 naive P14 cells [CD45.2+]) were vaccinated with LPS-matured, LCMV gp33-41 (gp33) peptide-pulsed dendritic cells (DC33) without or in combination with CpG (i.p.) (DC33/CpG). (B) Abundance of P14 cells (CD45.2+) in the spleen 7 days post-vaccination of P14 chimeric mice with DC33 or DC33/CpG. (C) Histogram overlays show Tcf1 expression in P14 cells (black open) relative to host CD8 T cells (gray open) and to no antibody control (gray fill). Numbers and bar graph depict the fraction of Tcf1low P14 cells. (D) IL-2 and IFN-γ production by P14 cells following in vitro restimulation with gp33 peptide. The bar graphs depict the percentage of IL-2+ (left) and IFN-γ+ (right) P14 cells. (E) Histogram overlays show granzyme B (GzmB) expression in P14 cells (black open) compared to isotype control (gray fill). Numbers indicate mean fluorescence intensity (MFI) of staining. The bar graph depicts the MFI of GzmB staining whereby the background was subtracted. (F) CD127 versus KLRG1 expression among gated P14 cells. The bar graphs indicate the percentage of CD127− KLRG1+ SLEC (left) and CD127+ KLRG1− MPEC P14 cells (right). (G and H) Histogram overlays show Tcf1 expression (G) or GzmB expression (H) in gated CD127+ KLRG1− (black open) and CD127− KLRG1+ P14 cells (gray open) compared to no antibody control (gray fill) and host CD8 T cells (dotted border, gray fill). Numbers depict the fraction of Tcf1low P14 cells (G) or the MFI of GzmB staining (H). All bar graphs show means (±SD, n = 5). Data are representative of 2 or more independent experiments. Statistical significance was determined using unpaired t tests (∗∗∗p < 0.001; ∗∗p < 0.01; ∗p < 0.05; ns, not significant [p > 0.05]). See also Figure S1. Cell Reports 2018 22, 2107-2117DOI: (10.1016/j.celrep.2018.01.072) Copyright © 2018 The Author(s) Terms and Conditions

Figure 2 IL-12 (p40) Blockade Prevents Inflammation-Induced Tcf1 Downregulation P14 chimeric mice were vaccinated with DC33/CpG and treated with isotype control antibody (Ab) or anti-IL-12 (p40) Ab. (A) Tcf1 expression in P14 cells (black open) compared to host CD8 T cells (gray open) and no antibody control (gray fill). Numbers and bar graph depict the percentage of Tcf1low cells in P14 cells. (B) The bar graph depicts the abundance of P14 cells in the spleen. (C) Production of IL-2 and IFN-γ following in vitro restimulation. The bar graphs depict the percentage of IL-2+ (left) and IFN-γ+ (right) P14 cells. (D) GzmB expression in P14 cells (black open) compared to isotype control (gray fill). The bar graph depicts the MFI of GzmB staining whereby background was subtracted. (E) CD127 versus KLRG1 expression among gated P14 cells. The bar graphs show the percentage of CD127− KLRG1+ SLECs (left) and CD127+ KLRG1− MPECs (right). Bar graphs show means (±SD, n = 5). Data are representative of at least 2 independent experiments. Statistical significance was determined using unpaired t tests (∗∗∗p < 0.001; ∗∗p < 0.01; ∗p < 0.05; ns, not significant [p > 0.05]). See also Figure S2. Cell Reports 2018 22, 2107-2117DOI: (10.1016/j.celrep.2018.01.072) Copyright © 2018 The Author(s) Terms and Conditions

Figure 3 Absence of IL12Rβ2 or STAT4 from CD8 T Cells Prevents Tcf1 Downregulation during Vaccination and Bacterial Infection (A–H) WT, IL12Rβ2−/− and STAT4−/− P14 chimeric mice were vaccinated with DC33 or DC33/CpG (A–D) or infected with L.m.-gp33 (E–H). (A and E) Tcf1 expression in P14 cells (black open) compared to host CD8 T cells (gray open) and Tcf7−/− P14 cells (control) (gray fill). The bar graph depicts the percentage of Tcf1low P14 cells. (B and F) IL-2 and IFN-γ production by P14 cells following in vitro restimulation with gp33 peptide. The bar graphs depict the percentage of IL-2+ (left) and IFN-γ+ (right) in P14 cells. (C and G) GzmB expression in P14 cells (black open) compared to isotype control (gray fill). The bar graph depicts the MFI of GzmB staining whereby background was subtracted. (D and H) CD127 versus KLRG1 expression among gated P14 cells. The bar graphs indicate the percentage of CD127− KLRG1+ SLECs (left) and CD127+ KLRG1− MPECs (right). Bar graphs depict means (±SD, n = 5). Data are representative of at least 2 independent experiments. Statistical significance was determined with one-way ANOVA tests (∗∗∗p < 0.001; ∗∗p < 0.01; ∗p < 0.05; ns, not significant [p > 0.05]). See also Figure S3. Cell Reports 2018 22, 2107-2117DOI: (10.1016/j.celrep.2018.01.072) Copyright © 2018 The Author(s) Terms and Conditions

Figure 4 IL-12 Represses Tcf7 Transcription and Establishes Asymmetric Tcf1 Abundance during Cell Division (A–C) Carboxyfluorescein succinimidyl ester (CFSE)-labeled P14 cells (CD45.2+) were activated in vitro with gp33 peptide-pulsed splenocytes (CD45.1+ or CD45.1+/2+). Representative flow cytometry plots show CFSE dilution versus Tcf1 expression at the indicated time point post stimulation in the absence of IL-12 (A), in the presence of IL-12 (B), and upon addition of IL-12 and the cell-cycle inhibitor aphidicolin at 48 hr post-stimulation (C). Numbers indicate the fraction cells in the respective quadrants. Note that CFSE was gated according to residual CFSE expression observed at 48 hr post-stimulation. The data shown are representative of 3 independent experiments. (D) P14 cells were activated as above for 72 hr, and Tcf7 mRNA levels were determined using qRT-PCR analysis. Data shown represent quadruplicate determinations and are representative of 2 independent experiments. (E) P14 cells were activated as above, and the DNA methyltransferase inhibitor decitabine (black border, gray fill) or vehicle (DMSO) (gray border, open) was added at 48 hr post stimulation. Tcf1 expression was measured at 72 hr post-stimulation. The bar graph depicts the mean MFI (±SD) of Tcf1 expression normalized to P14 cells activated with gp33 + IL-12 (100%) from n = 4 (vehicle) or n = 8 (decitabine) independent determinations. Statistical significance was determined using an unpaired t test (∗∗∗p < 0.001; ∗p < 0.05; ns, not significant [p > 0.05]). See also Figure S3. Cell Reports 2018 22, 2107-2117DOI: (10.1016/j.celrep.2018.01.072) Copyright © 2018 The Author(s) Terms and Conditions

Figure 5 Tcf7−/− CD8 T Cells Undergo Effector Differentiation Even in the Absence of Systemic Inflammation WT or Tcf7−/− P14 chimeric mice were vaccinated with DC33 or DC33/CpG. (A) Abundance of P14 cells (CD45.2+) in the spleen at day 7 post-vaccination. (B) IL-2 and IFN-γ production by P14 cells following in vitro restimulation with gp33 peptide. The bar graphs show the percentage of P14 cells producing IL-2 (left) and IFN-γ (right). (C) GzmB expression (black open) compared to isotype control (gray fill). The bar graph shows the MFI of GzmB staining whereby background was subtracted. (D) CD127 versus KLRG1 expression among gated P14 cells. The bar graphs show the percentage of CD127− KLRG1+ SLECs (left) and CD127+ KLRG1− MPECs (right). (E) Tcf7lox/lox R26-lox-YFP P14 cells (CD45.2+) were treated with Tat-Cre in vitro, and the frequency of YFP+ P14 cells (Tcf7 deleted) was determined after 48 hr of culture in vitro (left). Histogram overlay shows Tcf1 expression by flow-sorted YFP+ (gray fill) and YFP− P14 cells (black open) (right). Numbers indicate the percentage of cells in the respective quadrants (left) and the percentage of Tcf1hi cells (right). (F) Tat-Cre-treated Tcf7lox/lox R26-lox-YFP P14 cells (CD45.2+) were transferred into CD45.1+ mice, which were then vaccinated with DC33. The frequency of YFP+ P14 cells (Tcf7 deleted) was determined 7 days later (left). Tcf1 expression by flow sorted YFP+ (gray fill) and YFP− P14 cells (black open) (right). Numbers indicate the percentage of cells in the respective quadrants (left) and the percentage of Tcf1hi cells (right). (G and H) Gated YFP+ (Tcf7-deleted) and YFP− P14 cells (non-deleted) were analyzed for the percentage of cells producing IL-2 (G) and the percentage of CD127− KLRG1+ SLECs and CD127+ KLRG1− MPECs (H). Bar graphs depict means (±SD, n = 5). Data are representative of at least two independent experiments. Statistical significance was determined with unpaired t tests (∗∗∗p < 0.001; ∗∗p < 0.01; ∗p < 0.05; ns, not significant [p > 0.05]). See also Figure S4. Cell Reports 2018 22, 2107-2117DOI: (10.1016/j.celrep.2018.01.072) Copyright © 2018 The Author(s) Terms and Conditions

Figure 6 Tcf1 Is Essential for Default Central Memory Formation in Response to DC33 Vaccination WT or Tcf7−/− P14 chimeric mice were vaccinated with DC33. (A) Abundance of P14 cells in the spleen 40 days post-vaccination. (B) CD127 versus CD62L expression among gated P14 cells. The bar graphs show the percentage of CD127+ CD62L+ (central memory phenotype) cells. (C) IL-2 and IFN-γ production by P14 cell following in vitro restimulation. The bar graphs depict the percentage of IL-2+ (left) and IFN-γ+ (right) P14 cells. (D) GzmB expression in P14 cells (black open) compared to isotype control (gray fill). Numbers indicate the MFI of staining, and the bar graph depicts the MFI of GzmB staining whereby background was subtracted. (E) WT and Tcf7−/− P14 memory cells (CD45.2+) were flow sorted, and equal numbers (104 cells) were transferred into secondary recipients (CD45.1+), which were infected with LCMV cl13. The abundance of P14 cells was determined 8 days later. The bar graph shows fold expansion of P14 cells compared to input, assuming 10% take of input cells. Bar graphs depict means (±SD, n = 3–5). Data are representative of two independent experiments. Statistical significance was determined with unpaired t tests (∗∗∗p < 0.001; ∗∗p < 0.01; ∗p < 0.05; ns, not significant [p > 0.05]). See also Figure S4. Cell Reports 2018 22, 2107-2117DOI: (10.1016/j.celrep.2018.01.072) Copyright © 2018 The Author(s) Terms and Conditions

Figure 7 Memory Precursor Cells Lacking Tcf1 Overexpress IL12Rβ2 and Transcription Factors Promoting Effector Differentiation (A–C) WT or Tcf7−/− P14 chimeric mice were vaccinated with DC33. 7 days later, P14 MPECs and SLECs, as well as naive P14 cells, were flow sorted and analyzed for the expression of IL12Rβ2−/− (A), Bcl6 and Eomes (B), and Prdm1 (Blimp1), Tbx21 (T-bet), Id2, and Zeb2 (C). Gene expression is shown relative to HPRT. Bar graphs depict means (±SD) from quadruplicate determinations. Data are from a single experiment. Equivalent results were obtained using DC33/CpG vaccination. Statistical significance was determined with unpaired t tests (∗∗∗p < 0.001; ∗∗p < 0.01; ∗p < 0.05; ns, not significant [p > 0.05]). See also Figure S5. Cell Reports 2018 22, 2107-2117DOI: (10.1016/j.celrep.2018.01.072) Copyright © 2018 The Author(s) Terms and Conditions